Entrance: Online Customer Service -> Smart Customer Service, click to open

Similar Question

Sometimes there may be differences in the description between the user's question and the preset question-answer pair maintained in the system. In order to enhance the accuracy of intelligent customer service for question identification as much as possible, we have added similar question methods. Users maintain the question-answer pair in the knowledge base. Data, multiple similar query methods can be maintained at the same time, for example, the main title can be maintained as "product price", and similar query methods can be maintained as "product quotation", "product price" and so on. identification can be made.

The FAQ data quoted by the smart customer service comes from the knowledge base application. Therefore, if you want to use the related functions of the smart customer service normally, you need to maintain the FAQ data in the knowledge base in advance, and bind the FAQ data under the smart customer service setting item, and the smart customer service can target The bound FAQ data range automatically queries questions and issues answers. The specific maintenance process can refer to the following figure

It is recommended that enterprises maintain exclusive knowledge categories for smart customer service

Create special scenes for smart customer service and bind repository classification data

The configuration information of the basic items of smart customer service is as follows

In addition to setting required items, it is recommended to sort out common customer questions based on the actual business of the enterprise and maintain them in recommended questions, which can greatly improve the practicability of smart customer service.

- The smart customer service is bound to the FAQ as follows, and the pre-maintained scene is bound to automatically fill in the knowledge classification.

Recommended Question

Usually, a question cannot completely solve the customer's problem. It is often necessary to design the hierarchical structure of the question, guide it step by step, and finally complete the service. This requires intelligent customer service to have the ability to continuously recommend questions. Enterprises can maintain FAQ question and answer data , bind the recommendation question, and after the intelligent customer service identifies the specific question and answers it, it can continue to recommend the question.

Proactively recommend questions and guide customers to ask precise questions.

After replying to precise questions, continue to recommend and guide questions based on the preset associated knowledge.

Questions sent by users can be split by keywords, and after the questions are matched, recommendations can be made based on the confirmation of the questions.

For multiple unidentified customer sessions, a manual transfer link will be prompted.